What if your drone or robot could operate in the field up to 40% longer?

Today’s Conventional Artificial Intelligence with its large language models runs on matrix mathematics and uses GPUs which were created for gaming. Conventional AI requires significant compute resources.

Dynamic neural model learning requires significantly less energy and resources in most applications 6000X less energy, 21,000X less resources for autonomy and delivers a 11% to 60% increase in battery range for a robot, or drone in autonomous mode. 

Key Takeaways from this session on making AI work in the real-world:

• An innovative way to map natural neural networks to a computing fabric (CPU, FPGA, ASIC).
• Dynamic Neural Models are compact and ‘live on the edge.’
• Extreme compute capability, very low power required.
• Dynamic neural models provide autonomy to robots and drones.
• Efficiencies increase flight time (battery) from 10% to 40%.